Flavor Evolution of the Neutronization Neutrino Burst from an O-Ne-Mg Core-Collapse Supernova

INT PUB 07-23
LA-UR-07-6726

Flavor Evolution of the Neutronization Neutrino Burst from an O-Ne-Mg

Core-Collapse Supernova

Huaiyu Duan,^ George M. Fuller,^' "'^ J. Carlson,'^ and Yong-Zhong Qian^

'institute for Nuclear Theory, University of Washington, Seattle, WA 98195
'^Department of Physics, University of California, San Diego, La Jolla, CA 92093
'^Theoretical Division, Los Alamos National Laboratory, Los Alamos, NM 87545
'^School of Physics and Astronomy, University of Minnesota, Minneapolis, MN 55455

We present results of 3-neutrino flavor evolution simulations for the neutronization burst from an
O-Ne-Mg core-collapse supernova. We find that nonlinear neutrino self-coupling engineers a single
spectral feature of stepwise conversion in the inverted neutrino mass hierarchy case and in the normal
mass hierarchy case, a superposition of two such features corresponding to the vacuum neutrino
mass-squared differences associated with solar and atmospheric neutrino oscillations. These neutrino
spectral features offer a unique potential probe of the conditions in the supernova environment and
may allow us to distinguish between O-Ne-Mg and Fe core-collapse supernovae.

PACS numbers: 14.60.Pq, 97.60.Bw

In this Letter we suggest an exciting new neutrino
signal-based probe of conditions deep inside a supernova.
', We do this by performing the first fully self-coupled 3-
neutrino flavor (3 x 3) evolution calculations. Stars of
', ^ 8-11 Mq develop degenerate O-Ne-Mg cores, at least
some of which eventually collapse to produce supernovae
[ (e.g., Refs. [H, 0, Q)- The matter density falls off so
steeply in the region between such a core and the hydro-
I gen envelope that there is little hindrance to the outgoing
supernova shock. Consequently, O-Ne-Mg core-collapse
supernovae are the only case where neutrino-driven ex-
plosion has been demonstrated by several groups 0, 0, 0] .
Such supernovae may be the site for producing the heav-
iest elements by rapid neutron capture and may also
] explain the observed subluminous supernovae They
are expected to be relatively common because the known
] progenitors of most core-collapse supernovae lie in the
■ mass range ~ 8-20 Mq (e.g., Ref. Q).

The region of steeply-falling matter density immedi-
ately above an O-Ne-Mg core provides an extremely in-
teresting environment for studying neutrino flavor evolu-
tion. For the vacuum neutrino mass-squared differences
^"^atm ^^'^ Attiq associated with atmospheric and solar
neutrino oscillations, respectively, the two correspond-
ing conventional Mikheyev-Smirnov-Wolfenstein (MSW)
0, [lot resonances occur with very small radial separa-
tion in this region. As the neutrino number density de-
creases much more gently with radius than the matter
density, neutrino self-coupling can affect flavor evolution
associated with both Aml^^ and Atoq , and a full treat-
ment of 3 X 3 mixing appears to be required. To identify
clearly any new physics, we study the relatively simple
case of the neutronization burst, which consists of pre-
dominantly Ve emitted when the shock breaks through
the neutrino sphere.

Traditional analyses of flavor evolution of supernova
neutrinos are based on the pure matter-driven MSW ef-
fect (see, e.g., Refs. [HI [H [13, El). The evolution of

neutrino flavor state in matter is described by the
Schrodinger-like equation.

(1)

where t is an AfRne parameter along the neutrino
worldline, and the Hamiltonian H is composed of two
pieces: H — Hva.c + -f^matt- The matter contribution is
{vaWraM^p) = ^/^G-pUeSapSea, whcrc Gp IS the Fcrmi
constant, Ue is the electron number density, and \va) de-
notes a pure flavor state with a = e, /x, r. The vacuum
piece of H is {va\Hva.c\vp) = {2E^)~^{UMW)af3, where
El, is the neutrino energy. The transformation Uai re-
lates pure flavor state [vn) to vacuum mass eigenstate
l^i) (see Chap. 13 of Ref. [15| for our convention): —

The mass matrix is diagonal in the

2 2

mf-rrij.

vacuum mass basis, M = diag(0, Ani^i, Am^i + Am|
where the mass-squared differences are Amf^
In calculations presented here we take the three mix
ing angles and the CP violating phase to be 9i2 = 0.6,
^23 — 7r/4, — 0.1, 5 — 0, respectively. We take

8 X 10-^ eV^

Atoq and Am§2 = ±3 x

10~ eV ~ ±Ato^j„-j, where the plus (minus) sign is for
the normal (inverted) mass hierarchy.

In pure matter-driven MSW evolution, for small 6*13,
the Ve survival probability Pu^u,, — |(^^e|V')P can be fac-
torized P^^u, = Pu.u.Pku., where P^^^^ and P^^^
are the survival probabilities in 2-flavor (2x2) mixing
processes at the Aml^^ and Atoq scales, respectively. In
other words, the full 3x3 MSW result is the superposition
of two independent 2x2 MSW scenarios, one for each of
the solar and atmospheric mass-squared differences.

Using the profile for the O-Ne-Mg core model of
Refs. P, Q and the neutrino mixing parameters given
above, we show P^^^^ as a function of E,, in Fig. [1] as-
suming pure matter-driven MSW evolution. The results
shown are for radius r = 5000 km, where the vacuum
Hamiltonian dominates for most neutrino energies. The

2

20

E„ (McV)

FIG. 1: (Color online) Neutrino survival probabilities Pv^u^ as
functions of neutrino energy Eu for pure matter-driven MSW
evolution. The 2x2 flavor mixing cases with Am^ ~ Amatm
and ArriQ are shown as the dashed and dotted lines, respec-
tively. The 2x2 flavor mixing case with Am? ~ — Amatm
(not shown) corresponds to Pv^v^ — 1 for all energies. The
solid line {Ev) for full 3x3 flavor mixing with the

normal mass hierarchy. The 3x3 inverted mass hierarchy
case (not shown) is almost identical to the dotted line.

dashed and dotted lines in this figure show the 2x2
flavor mixing cases with the normal mass hierarchy for
Am^ = 3 X 10^3 (~ Ato^^^) and 8 x 10"^ eV^
(~ Amg), respectively. In these cases we take the ef-
fective 2x2 vacuum mixing angles to be = 0.1 and
0.6, respectively. We note that in either case MSW fla-
vor transformation for neutronization burst neutrinos of
average energy {E^^ ) = 1 1 MeV is neither fully adiabatic
[PvcVc — sin^ 6*) nor fully non-adiabatic [Pv^vc ~ cos^ &)
due to the rapid decrease of matter density p with ra-
dius in the region of interest (|d(ln/9)/dr| > 0.04km~^).
The spike in P^^j^^) (dashed fine) at E'^, ~ 8 MeV is
caused by a sharp change in rig at the base of the hy-
drogen envelope, where the electron fraction Y,, jumps
from 0.5 to ^ 0.85. The 2x2 inverted mass hierarchy
case with Am^ ~ — Am^^^ has Py^y^ ~ 1 for all energies
(i.e., no MSW resonance). In the complete 3x3 mixing
case with the normal mass hierarchy, Py^^^ is given by
the solid line. This case corresponds closely to a succes-
sion of two independent 2x2 mixing schemes, with the
solid line being approximately the product of the values
of the dashed (P^^^^) and dotted (F^^^J Hnes. The 3x3
inverted mass hierarchy case gives P^^vc nearly identical
to the dotted fine.

In supernovae, where neutrino luminosities are large,
neutrino- neutrino forward scattering contributes another
term for the Hamiltonian

[A [13, El

il,, = V2Gf

(2)

where sums over all background neutrino states ItAa)
with number density ny^\. To simplify the problem, we
adopt the "single-angle approximation" in which neutri-

nos emitted in all directions from the neutrino sphere
have the same flavor evolution histories as those with the
same energies but propagating along a radial trajectory.
With this approximation we have

E

A

D{r/Ru) L
2tiRI {E,

(3)

where -D(^) — — -^/l — ^^•^)^. In our calculations for
the neutronization burst we assume Ve is the only neu-
trino species emitted from the neutrino sphere (at radius
Ru — 60 km) and take the luminosity to be L^^ = 10^'^
erg/s. The Ve energy distribution function /^^{Ey) is
taken to be of Fermi-Dirac form with degeneracy param-
eter ?7 = 3 and with an average i^e energy (E^^) — 11
MeV. Full 2x2 multi-angle simulations show that the
single-angle approximation appears to be adequate for
qualitative studies of the collective flavor transformation
phenomena of interest here 1^, 2^, 21 1 .

Fig. [2] shows the results of single-angle simulations of
full 3x3 neutrino flavor evolution including nonlinear
neutrino self-coupling for the neutrino mixing and emis-
sion parameters given above. Results for both the in-
verted (upper panels) and normal (lower panels) neutrino
mass hierarchies are presented, again at radius r — 5000
km as in Fig. [TJ The left-hand panels show the proba-
bility |aj^. p = for neutrinos to be in each of the
mass eigenstates and the right-hand panels show the
probability lajy^P = for neutrinos to be in each
of the flavor states \i^a)-

The inverted neutrino mass hierarchy produces a step-
wise i'2/'^i conversion at energy E^ ~ 11 MeV [Fig.[5]Ja)].
This spectral swap feature can be understood in a 2 x 2
mixing scheme with Am? ~ AttIq (see, e.g., Ref. ^2^).
In this scheme the flavor evolution of a neutrino can
be represented as the precession of a spin or polariza-
tion vector in flavor isospace, in analogy to a magnetic
spin (e.g., Ref. [H). When neutrino number fluxes
are large, the neutrino self-coupling is strong and the
"magnetic spins" representing neutrinos can rotate col-
lectively in the region where a neutrino with a represen-
tative energy would experience a resonance in the pure
matter-drive MSW evolution [2^. This corresponds to
a neutrino-background-enhanced MSW-like flavor trans-
formation 23, 23, 24, 2^. Subsequently, the "magnetic
spins" will enter a collective precession mode. As neu-
trino fluxes become small at large radii and the collective
precession mode dies out, a mass- basis spectral swap is
established The swap point in the neutrino en-

ergy spectrum is determined by conservation of a mass-

1. In fact, the result of

basis "lepton number" [27|, [2

the full 3x3 calculation agrees very well with that of
the 2x2 calculation with Am^ ~ Amg. In contrast to
the pure-matter driven MSW evolution, neutrinos on the
two sides of the swap point appear to have experienced
almost fully adiabatic or fully non-adiabatic flavor trans-

3

FIG. 2: (Color online) Probabilities as functions of neutrino energy Ei, for neutrinos to be in each vacuum mass eigenstate
(|ai,. 1^, left panels) and flavor eigenstate (|a.yQp, right panels), respectively. The solid, dashed and dotted lines represent the
ui, !/2, and 1^3 states in the left panels and the u^, i/^, and Vt states in the right panels. Top (bottom) panels show the inverted
(normal) mass hierarchy case.

formation with those in the low (high) mass eigenstate
ending up on the right (left) hand side of the swap point.
We note that no neutrino-background-enhanced MSW-
like flavor transformation occurs at the Am^^^ scale in
the inverted mass hierarchy case. This is analogous to the
pure matter-driven MSW evolution. We also note that
conservation of the mass-basis lepton number prohibits
the formation of a spectral swap in the corresponding
2x2 mixing scheme with Am^ ~ — Am^tm because all
neutrinos start as Ve in our calculation [38[.

Because 9i2 is large, Pv,v^{Eu) exhibit a large oscil-
latory feature in the transition region near the stepwise
^^2/1^1 conversion point at E,y ~ 11 MeV [see Fig. (^fb)].
Outside this region, spectral swap can also be seen for
the neutrino flavor states. For example, Pu^v^ — |C^e2p —
0.32 (|;7ei|2 ~ 0.67) for < ^ MeV {eI > 16 MeV).
Note that |ai,^J^ is larger (smaller) than \au^\'^ in the en-
ergy regime > 13 MeV {E^ < 10 MeV). This is a
consequence of setting the CP-violating phase to <5 = 0.
As S is increased, \a^^\'^ increases (decreases) for E,y > 13
MeV (E^ < 10 MeV), and \a^J^ = \a^^\^ for 6 = 7r/2.
For S — TT, the \a^^ p and \a^^ p curves in Fig.[2Ib) switch
places.

Fig- EUc) shows that the normal neutrino mass hierar-
chy produces a superposition of two spectral swap fea-

tures, reminiscent of the factorization property of pure
matter-driven MSW evolution. The v^^jvi swap at Ey ~
12.7 MeV and the 1^2 /i^i swap at ii^^ ~ 15 MeV corre-
spond to those in the 2x2 schemes with Am^ ~ Aw?^^^
and Atoq, respectively. This result seems to justify
the 2x2 approximation used in previous work (e.g.,
Refs.di, S [in, m, m m, [13, [111 [33), but is some-
what surprising given the nonlinear nature of neutrino
self-coupling and the fact that the regions of collective
flavor transformation for Am^^j^ and Awq overlap with
each other. We note that the 1^3/1^2 swap is much sharper
than the vijvi swap. We also note that the dip (bump)
in la^gp (|ai/2p) centered at ~ 5.2 MeV corresponds
to the abrupt change in rig at the base of the hydrogen
envelope. This feature in the profile reduces the ef-
ficiency of the neutrino-background-enhanced MSW-like
transformation of Vf^ at the Am^^j^ scale. A similar fea-
ture is also present in the pure matter-driven MSW evo-
lution, but in a different energy range (see Fig.[T|).

As in the inverted mass hierarchy case, the spectral
swaps are also present in the flavor basis [see Fig. Hfd)].
Except for the moderate bump centered at Ey ~ 5.2
MeV, nearly all j/g's are transformed (with Py^y^ ~
[J/esI^ ~ 0.01) below the v^^jvi swap energy Ey ~ 12.7
MeV. In contrast, there is less significant v^, depletion

4

above the V2lv\ swap energy i^i, ~ 15 MeV, for which
energy range Py^v^ — |C^eiP — 0.67 would be expected.
The bump in Py^y^{Ey) at ~ 5.2 MeV corresponds

to the feature in

at the same energy as explained

above. On the other hand, we note that the apparent
peak of Pi,^i,^{Ey) in Ei, ~ 15.5 MeV is part of the same
oscillation feature discussed above for Fig. [2{b). This
oscillation feature will disappear at very large distances
where coherence has been lost.

The spectral swap features illustrated here for the neu-
tronization burst from an 0-Ne-Mg core-collapse super-
nova are not expected to be present for an Fe-core col-
lapse supernova. This is because at the neutronization
burst epoch, there is an extended region of high Ue above
an Fe core. So in this case high neutrino fluxes are al-
ways accompanied by high rig, which inhibits neutrino-
background-enhanced MSW-like flavor transformation.
The studies of neutrinos from Fe core-collapse supernovae
based on pure matter-driven MSW evolution show that
the Ve survival probability is Pu^u^ — sin^ 9q ~ 0.32
or less for either mass hierarchy (see, e.g.. Table I in
Ref. [33]). If enough high-energy {E^, > 15 MeV) neu-
trino events are collected from the neutronization burst
of a future Galactic supernova in both the charged-
current and neutral-current channels, the progenitor may
be identified as having an 0-Ne-Mg or Fe core based on
whether or not is the dominant species in the burst.
Because the total neutrino fluence over the ~ 10 ms du-
ration of the neutronization burst is only a small fraction
of that emitted during the first several seconds after the
onset of core collapse, collection of the required number
of events from the neutronization burst may be beyond
the capabilities of existing detectors but could be within
the reach of proposed megaton water Cherenkov detec-
tors [331, ISJ, |35| and liquid argon detectors like ICARUS

[3^, |37[. The low-energy (< 10 MeV) neutrino signals
in the 0-Ne-Mg core-collapse supernova neutronization
burst, though even more difficult to detect, carry infor-
mation that potentially can distinguish between the neu-
trino mass hierarchies [see Figs.[2][b) and (d)].

We thank K. Nomoto for providing the electron num-
ber density profile in his 0-Ne-Mg core model. We appre-
ciate the hospitality of the Institute for Nuclear Theory
at the University of Washington during the completion
of this work. This work was supported in part by NSF
grant PHY-04-00359 at UCSD, DOE grants DE-FG02-
87ER40328 at UMN, DE-FG02-00ER41132 at INT, an
IGPP/LANL mini-grant, and by the DOE Office of Nu-
clear Physics, the LDRD Program and Open Supercom-
puting at LANL.

[1] K. Nomoto, Astrophys. J. 277, 791 (1984).
[2] K. Nomoto, Astrophys. J. 322, 206 (1987).

[3] C. Ritossa, E. Garci'a-Berro, and I. J. Iben, Astrophys.

J. 515, 381 (1999).
[4] R. Mayle and J. R. Wilson, Astrophys. J. 334, 909

(1988).

[5] F. S. Kitaura, H.-T. Janka, and W. Hillebrandt, Astron.

Astrophys. 450, 345 (2006), astro-ph/05 12065.
[6] L. Dessart et al., Astrophys. J. 644, 1063 (2006), astro-

ph/0601603.

[7] H. Ning, Y. Z. Qian, and B. S. Meyer, Astrophys. J. 667,

L159 (2007), arXiv:0708.1748 [astro-ph].
[8] W.-D. Li et al., Astrophys. J. 641, 1060 (2006), astro-

ph/0507394.

[9] L. Wolfenstein, Phys. Rev. D17, 2369 (1978).
[10] S. P. Mikheyev and A. Y. Smirnov, Yad. Fiz. 42, 1441

(1985), [Sov. J. Nucl. Phys. 42, 913 (1985)].
[11] T.-K. Kuo and J. T. Pantaleone, Phys. Rev. D37, 298

(1988).

[12] A. S. Dighe and A. Y. Smirnov, Phys. Rev. D62, 033007

(2000), hep-ph/9907423.
[13] C. Lunardini and A. Y. Smirnov, JCAP 0306, 009

(2003), hep-ph/0302033.
[14] J. P. Kneller and G. C. McLaughlin, Phys. Rev. D73,

056003 (2006), hep-ph/0509356.
[15] W.-M. Yao et al., J. Phys. G33, 1 (2006).
[16] G. M. FuUer, R. W. Mayle, J. R. Wilson, and D. N.

Schramm, Astrophys. J. 322, 795 (1987).
[17] J. T. Pantaleone, Phys. Rev. D46, 510 (1992).
[18] G. Sigl and G. Raffelt, Nucl. Phys. B406, 423 (1993).
[19] H. Duan, G. M. Fuller, J. Carlson, and Y.-Z. Qian, Phys.

Rev. D74, 105014 (2006), astro-ph/0606616.
[20] A. Esteban-Pretel, S. Pastor, R. Tomas, G. G. Raf-
felt, and G. Sigl, Phys. Rev. D76, 125018 (2007),

arXiv:0706.2498 [astro-ph].
[21] G. L. Fogli, E. Lisi, A. Marrone, and A. Mirizzi (2007),

arXiv:0707.1998 [hep-ph].
[22] H. Duan, G. M. Fuller, and Y.-Z. Qian, Phys. Rev. D76,

085013 (2007), arXiv:0706.4293 [astro-ph].
[23] S. Pastor and G. Raffelt, Phys. Rev. Lett. 89, 191101

(2002), astro-ph/0207281.
[24] Y. Z. Qian and G. M. Fuller, Phys. Rev. D51, 1479

(1995), astro-ph/9406073.
[25] G. M. Fuller and Y.-Z. Qian, Phys. Rev. D73, 023004

(2006), astro-ph/0505240.
[26] H. Duan, G. M. Fuller, J. Carlson, and Y.-Z. Qian, Phys.

Rev. D75, 125005 (2007), astro-ph/0703776.
[27] G. G. Raffelt and A. Y. Smirnov, Phys. Rev. D76,

081301 (R) (2007), arXiv:0705.1830 [hep-ph].
[28] S. Hannestad, G. G. Raffeh, G. Sigl, and Y. Y. Y. Wong,

Phys. Rev. D74, 105010 (2006), astro-ph/0608695.
[29] A. B. Balantekin and H. Yiiksel, New J. Phys. 7, 51

(2005), astro-ph/0411159.
[30] H. Duan, G. M. Fuller, and Y.-Z. Qian, Phys. Rev. D74,

123004 (2006), astro-ph/0511275.
[31] H. Duan, G. M. Fuller, J. Carlson, and Y.-Z. Qian, Phys.

Rev. Lett. 97, 241101 (2006), astro-ph/0608050.
[32] H. Duan, G. M. Fuller, J. Carlson, and Y.-Z. Qian, Phys.

Rev. Lett. 99, 241802 (2007), arXiv:0707.0290 [astro-ph].
[33] M. Kachelriess et al, Phys. Rev. D71, 063003 (2005),

astro-ph/0412082.
[34] Y. Itow et al. (The T2K) (2001), hep-ex/0106019.
[35] M. Goodman et al., UNO whitepaper (2001),

PREPRINT-SBHEPOl-3.
[36] I. Gil-Botella and A. Rubbia, JCAP 0310, 009 (2003),

hep-ph/0307244.

5

[37] C. Vignoli, D. Barni, J. M. Disdior, D. Rampoldi, and [38] A. Mirizzi (2007), private communication.
G. Passardi (ICARUS), AIP Conf. Proc. 823, 1643
(2006).